metadata

license: apache-2.0
base_model: google/vit-base-patch16-224
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - f1
model-index:
  - name: finetuned-bin
    results: []

finetuned-bin

This model is a fine-tuned version of google/vit-base-patch16-224 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 2.7526
Accuracy: 0.0582
F1: 0.0356

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0002
train_batch_size: 16
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 4
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
2.7759	0.13	100	2.7526	0.0582	0.0356
2.7726	0.25	200	2.7794	0.0627	0.0208
2.7741	0.38	300	2.7754	0.0853	0.0450
2.775	0.5	400	2.7739	0.0818	0.0357
2.7738	0.63	500	2.7691	0.1204	0.0434
2.7737	0.75	600	2.7744	0.0573	0.0349
2.7732	0.88	700	2.7759	0.0484	0.0244
2.7746	1.0	800	2.7592	0.0942	0.0434
2.7737	1.13	900	2.7727	0.1098	0.0541
2.7715	1.25	1000	2.7719	0.0893	0.0414
2.7742	1.38	1100	2.7985	0.0	0.0
2.7715	1.51	1200	2.7729	0.024	0.0158
2.7698	1.63	1300	2.7711	0.0649	0.0344
2.7717	1.76	1400	2.7709	0.0858	0.0387
2.7708	1.88	1500	2.7726	0.0587	0.0368
2.7736	2.01	1600	2.8029	0.0	0.0
2.7726	2.13	1700	2.7743	0.088	0.0327
2.7734	2.26	1800	2.7734	0.0284	0.0163
2.7726	2.38	1900	2.7731	0.0578	0.0330
2.7733	2.51	2000	2.7711	0.1098	0.0498
2.7729	2.63	2100	2.7769	0.0018	0.0019
2.773	2.76	2200	2.7631	0.1076	0.0319
2.7743	2.89	2300	2.7768	0.0	0.0
2.7736	3.01	2400	2.7883	0.0	0.0
2.7726	3.14	2500	2.7727	0.104	0.0495
2.7727	3.26	2600	2.7694	0.0969	0.0569
2.7733	3.39	2700	2.7861	0.0004	0.0008
2.7739	3.51	2800	2.7727	0.1022	0.0575
2.7734	3.64	2900	2.7728	0.112	0.0478
2.7732	3.76	3000	2.7729	0.1236	0.0567
2.7736	3.89	3100	2.7727	0.1218	0.0587

Framework versions

Transformers 4.38.2
Pytorch 2.2.1+cu121
Datasets 2.18.0
Tokenizers 0.15.2